Molecular Phylogenetics and Evolution 106 (2017) 228–240
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Resolving incongruence: Species of hybrid origin in Columnea (Gesneriaceae) ⇑ James F. Smith a, , John L. Clark b,c, Marisol Amaya-Márquez d, Oscar H. Marín-Gómez d,e a Department of Biological Sciences, Boise State University, 1910 University Drive, Boise, ID 83725-1515, USA b Department of Biological Sciences, The University of Alabama, Box 870345, Tuscaloosa, AL 35487, USA c The Lawrenceville School: Science Department, The Lawrenceville School, 2500 Main Street, Lawrenceville, NJ 08648, USA d Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Apartado 7495, Bogotá, Colombia e Posgrado en Ciencias, Instituto de Ecología, INECOL A.C, Carretera antigua a Coatepec 351, El Haya, Xalapa 91070, Veracruz, Mexico article info abstract
Article history: Speciation by hybridization has long been recognized among plants and includes both homoploid and Received 12 August 2016 allopolyploid speciation. The numbers of presumed hybrid species averages close to 11% and tends to Revised 30 September 2016 be concentrated in a subset of angiosperm families. Recent advances in molecular methods have verified Accepted 3 October 2016 species of hybrid origin that had been presumed on the basis of morphology and have identified species Available online 5 October 2016 that were not initially considered hybrids. Identifying species of hybrid origin is often a challenge and typically based on intermediate morphology, or discrepancies between molecular datasets. Keywords: Discrepancies between data partitions may result from several factors including poor support, incom- Hybrid species plete lineage sorting, or hybridization. A phylogenetic analysis of species in Columnea (Gesneriaceae) Interspecific hybrids nrDNA capture indicated significant incongruencies between the cpDNA and nrDNA datasets. Tests that examined whether one or both of the datasets had the phylogenetic signal to reject the topology of the alternate dataset (Shimodaira and Hasegawa [SH] and approximately unbiased [AU] tests) indicated significant dif- ferences between the topologies. Splitstree analyses also showed that there was support for the place- ment of the discrepant taxa in both datasets and that the combined data placed the putative hybrid species in an intermediate position between the two datasets. The genealogical sorting index (GSI) implied that coalescence in nrDNA had occurred in all species where more than a single individual had been sampled, but the GSI value was lower for the cpDNA of most of the putative hybrids, implying that these regions have not yet coalesced in these lineages despite being haploid. The JML test that eval- uates simulated species pairwise distances against observed distances also implies that observed nrDNA data generate shorter distances than simulated data, implying hybridization. It is most likely that C. gigantifolia, C. rubriacuta, and C. sp. nov. represent a lineage from a hybrid ancestor, but C. moorei may be a more recent hybrid and may still be undergoing hybridization with sympatric species. Ó 2016 Elsevier Inc. All rights reserved.
1. Introduction and founder effects (Glor et al., 2005; Smith et al., 2014; Schulte et al., 2015). More rapid means of generating species include The process of speciation has long been a focus of evolutionary hybridization either through homoploid or polyploid means and and systematic biology (Darwin, 1859; Mayr, 1942; Levin, 1978, generally is estimated to be responsible for 11% of all species 2000; Coyne and Orr, 2004; Lexer and Widmer, 2008; Givnish, (Stebbins, 1959; Rieseberg et al., 1996a; Rieseberg, 1997, 2006; 2010, 2015; Sochar et al., 2015; Kadereit, 2015; Arnold, 2016). Wendel and Cronn, 2003; Soltis and Soltis, 2009; Abbott et al., Allopatric speciation is a well documented process and often is 2013). Detecting hybrids is not always a simple task. Intermediacy divided into vicariant speciation where landscapes change to iso- in morphology has long been a staple for detecting hybrids, but it is late populations from each other (Pereira and Baker, 2004; now clear that hybrids may possess transitive morphologies: char- Struwe et al., 2009; Bentley et al., 2014), or isolation by dispersal acter states that are unknown in the parental species (Rieseberg and Ellstrand, 1993). For example, Castilleja christii N. H. Holmgren was recently documented to be a homoploid hybrid species (Clay ⇑ Corresponding author. et al., 2012), but was never suspected to be of hybrid origin. In E-mail address: [email protected] (J.F. Smith). http://dx.doi.org/10.1016/j.ympev.2016.10.001 1055-7903/Ó 2016 Elsevier Inc. All rights reserved. J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 229 large part the lack of suspicion stemmed from C. christii having pri- Aeschynanthus (Denduangboripant and Cronk, 2000). Therefore marily yellow bracts in contrast to the crimson and scarlet bracts we evaluate whether incomplete lineage sorting or hybridization of the parental species, C. miniata and C. linariifolia. Castilleja christii may be most prevalent. We examined our data using several alter- was shown to be a hybrid only when direct sequencing of low copy native analytical approaches including using Shimodaira and Hase- nuclear genes demonstrated multiple peaks in chromatograms that gawa (SH; Shimodaira and Hasegawa, 1999) and approximately corresponded to the two separate peaks that were present in the unbiased (AU; Shimodaira, 2002) tests and two coalescent parental copies of the homologous gene (Clay et al., 2012). approaches. The first used the genealogical sorting index (GSI, Modern means of detecting hybrids often occur when a species Cummings et al., 2008) that evaluates whether a pre-defined clade is placed in different clades when chloroplast and nuclear DNA is monophyletic even when it may not be recovered as mono- data are analyzed independently (Smith and Sytsma, 1990; phyletic in standard phylogenetic analyses. Clades that are mono- Rieseberg and Soltis, 1991; Rieseberg et al., 1996a or b; Baum phyletic are more likely to have achieved coalescence and et al., 1998; Wendel and Doyle, 1998; Linder and Rieseberg, therefore are less likely to reflect incomplete lineage sorting 2004; Howarth and Baum, 2005; Friar et al., 2008; Rothfels et al., (Palumbi et al., 2001; Hedrick, 2007; de Viliers et al., 2013). 2015; Walker et al., 2015). Discrepancies between data partitions Joly et al. (2009) and Joly (2012) developed the software JML to have posed challenges to phylogenetic analyses since systematists detect whether intraspecific variation in a gene that produces a started comparing more than a single dataset (Kluge, 1989; Smith discrepant relationship relative to another gene is likely the result and Sytsma, 1994; Mason-Gamer and Kellogg, 1996; Smith, 2000). of incomplete coalescence. If incomplete lineage sorting can be Incongruences can occur from a multitude of causes and the great- eliminated, the probability that the incongruency is the result of est challenge is to resolve the source of the incongruency. One pos- hybridization is increased. The software uses a posterior distribu- sibility is that the incongruence is the result of poor support in one tion of species trees, simulates gene trees and DNA sequences, then or more of the partitions (Farris et al., 1994; Seelanan et al., 1997; calculates the minimum distance between simulated sequences for Morrison, 2009). As a result, random noise and homoplasy may all pairs of species. These are then compared to the empirical data. have as much influence on the resulting topology as the phyloge- If the observed distances are smaller than the simulated distances, netic signal in the data. Such discrepancies are largely overcome then hybridization is a better explanation to account for the more with more data (sometimes by combining several datasets with recent common ancestry of the sequences than incomplete lineage weak, but essentially congruent signal (Smith, 2000) or data that sorting. has a higher proportion of phylogenetically informative characters The taxonomic focus of the present study is the genus Columnea (Small et al., 1998). In other cases the datasets may each strongly (Gesneriaceae). Columnea is a genus of over 209 species that has support conflicting relationships. In such cases, the challenge is been the focus of several recent phylogenetic investigations to to discover the biological explanation for the incongruency and re-evaluate the subgeneric classification system and to explore requires knowledge beyond what is recovered in the phylogenetic character state evolution (Smith et al., 2013; Schulte et al., 2014, analyses. In cases where there is potential for paralogy such as low 2015). Preliminary analyses that included two individuals of C. copy nuclear genes, the inclusion of non-orthologous loci for some rubriacuta (Wiehler) L.P. Kvist & L.E. Skog, and one cultivated indi- taxa will generate incongruency (Rokas et al., 2003). In these cases vidual of C. moorei C.V. Morton, had indicated these species were removing the paralogs may provide a simple answer as long as incongruently placed in the phylogeny using cpDNA or nrDNA. resolving which sequences are the ones in conflict can be identified They were excluded from the analyses pending increased sampling (Schulte et al., 2015). In other cases, incongruencies can occur with or further analyses. Collections made in Colombia in 2013 afforded loci that are either haploid (mitochondrial or chloroplast DNA increased sampling of populations and individuals of C. rubriacuta regions) or are considered to be essentially single copy as the result and these additional samples resulted in similar incongruence that of concerted evolution such as nuclear ribosomal RNA regions had been seen with the two previously sampled individuals. Addi- (Álvarez and Wendel, 2003; Feliner and Rosselló, 2007). Here, the tionally, previously unsampled C. gigantifolia L.P. Kvist & L.E. Skog challenge is to tease out whether the incongruency is the result and an undescribed species from Colombia showed a similar pat- of incomplete lineage sorting (Avise et al., 1983; Pamilo and Nei, tern of incongruence to that of C. rubriacuta. Interspecific 1988; Doyle, 1992; Maddison, 1997; Rosenberg, 2002, 2003), or hybridization is well documented in Columnea (Moore, 1954; Lee hybridization (Alice et al., 2001; Martinsen et al., 2001; Lumaret and Sherk, 1963; Sherk and Lee, 1967; Morley, 1971, 1975, 1976; and Jabbour-Zahab, 2009; Pirie et al., 2009; Jabaily and Sytsma, Saylor, 1971; Byrne and Morley, 1976; Wiehler, 1976, 1983; 2010) because the patterns can be similar (Holder et al., 2001). Smith, 1991, 1994) and naturally occurring hybrids have been Resolving between incomplete lineage sorting and hybridiza- speculated on, but to date, have only been documented with mor- tion is not a trivial task and in many studies where discrepancies phological and cytological data between the Jamaican species C. are detected, authors often resort to ad hoc explanations about urbanii W.T. Stearn and C. rutilans Swartz (Morley, 1971) and could the potential geographic overlap that may bias or preclude not reject a hybrid origin of C. querceti Oerst. (Byrne and Morley, hybridization, and current population size or time since divergence 1976). Here we examine the potential hybrid origin of C. rubriacuta, from a common ancestor to bias or preclude incomplete lineage C. gigantifolia, C. sp. nov., and C. moorei using phylogenetic analyses sorting (Wendel and Doyle, 1998; Comes and Abbott, 2001; of DNA sequences and coalescent analyses to assess the degree to Jabaily and Sytsma, 2010). A growing number of studies to date which the different DNA sequence partitions have achieved coales- are attempting to analyze data to discriminate between these cence within each of the species. factors (Buckley et al., 2006; Kubatko and Degnan, 2007; Holland et al., 2008; Maureira-Butler et al., 2008; Joly et al., 2. Materials and methods 2009; Pirie et al., 2009; Polihronakis, 2009; Willyard et al., 2009; Pelser et al., 2010; de Viliers et al., 2013; Yu et al., 2011, 2013; 2.1. Taxon sampling, DNA extraction, amplification, and alignment Kuppler et al., 2015). In the present study, chloroplast DNA and nuclear ribosomal A complete list of samples, voucher specimens, and GenBank DNA gave supported conflicting placement for four species in numbers for all sequences used in all analyses is in Appendix A. Columnea (Gesneriaceae). Neither of these is likely to include par- The majority of species and accessions sampled is based on earlier alogs although it should be noted that evidence for incomplete phylogenetic analyses of the genus. Collections of C. gigantifolia concerted evolution in nrDNA has been documented in and C. sp. nov. were opportunistic collections made in 2013 in 230 J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240
Colombia. All available accessions of C. rubriacuta were sampled. all datasets. Maximum parsimony analyses were performed using These included numerous collections made in Colombia in 2013 PRAP2 (Müller, 2004; using the default settings but uploading and all available collections made by J. L. Clark including one the nexus file for each dataset) in conjunction with PAUP⁄ v4.0 accession of Columnea albovinosa (Freiberg) J.L. Clark & L.E. Skog b10 (Swofford, 2002). Bootstrap support (BS; Felsenstein, 1985) that is often annotated as a heterotypic synonym of C. rubriacuta. was estimated with 1000 heuristic replicates using PRAP2 Columnea moorei was sampled using one cultivated specimen and (Müller, 2004). Descriptive statistics reflecting the amount of phy- a recently collected sample from the wild. The recent collection logenetic signal in the parsimony analysis were given by the con- of C. moorei from the wild also afforded the opportunity to collect, sistency index (CI; Kluge and Farris, 1969), retention index (RI; and include all other species found in sympatry with C. moorei in Farris, 1989), and the resulting rescaled consistency index (RC; Panama. Farris, 1989). Maximum likelihood and ML bootstrapping was DNA was extracted from silica-dried leaf material using Qiagen investigated using RaxML-HPC2 (Stamatakis, 2006; Stamatakis DNeasy plant mini kits (Valencia, CA) according to manufacturer’s et al., 2008) on the CIPRES portal (Miller et al., 2010) using the instructions. The outgroup included four species of Glossoloma defaults, but allowing bootstrap replicates to be terminated Hanst. that has been identified as the sister genus to Columnea automatically. (Clark et al., 2012). Our sampling relied on species that were sam- Bayesian inference analyses were performed using optimal sub- pled in previous studies (Smith et al., 2013; Schulte et al., 2014, stitution models suggested by jModeltest 3.6 (Posada, 2008). The 2015) but reduced the number of individuals for species that had Akaike information criterion (AIC), which allows non-nested mod- previously been recovered as monophyletic. We also included C. els to be evaluated, was used as a selection criterion (Posada and gigantifolia (two individuals), C. sp. nov. (three individuals), C. Buckley, 2004) for all datasets. All analyses were run with four rubriacuta (29 individuals), and C. moorei (two individuals) chains, for ten million generations. Convergence was determined (Fig. 1). All samples were collected in the wild with the exception by viewing in Tracer v1.3 (Rambaut and Drummond, 2005), and of one accession of C. moorei (JLC 11307) that was made from cul- a burnin of 50,000 generations was discarded prior to sampling tivated material., These latter four species were included because the posterior distribution for all BI analyses. All BI analyses were preliminary analyses had indicated discrepancies in their phyloge- repeated twice to ensure that parameter estimates converged to netic relationships between cpDNA and nrDNA. similar values. The separate runs were compared using the online The following five cpDNA gene regions were chosen: trnQ-rps16 version of Are We There Yet (AWTY; Nylander et al., 2008)asa spacer (Shaw et al., 2007), rpl32-trnLUAG spacer (Shaw et al., 2007), means of determining if the separate chains approximated the rps16 intron (Oxelman et al., 1997), trnS-G spacer (Hamilton, 1999), same target distribution. and trnH-psbA spacer (Clark et al., 2006). The cpDNA gene regions were treated as a single partition because they are inherited as a 2.4. Estimating potential for hybridization single non-recombining unit. The two nuclear DNA gene regions, ITS (Baldwin et al., 1995) and ETS (Baldwin and Markos, 1998) Initially we viewed the cpDNA, nrDNA and combined datasets were treated as a single partition. These datasets were analyzed using Splitstree (Huson, 1998; Morrison, 2009) to get an estimate separately as the cpDNA and nrDNA data, respectively, and as a of the support that was present in each data set. Network relation- concatenated dataset that included all regions for all samples. ships for each of the datasets (cpDNA, nrDNA, concatenated) were We were investigating the potential for hybrids in this study, examined in Splitstree using neighbornet optimization. therefore we also included analyses that had all cpDNA or all The SH and AU tests were performed with the same constraints. nrDNA for the individuals that were discrepant between the two Columnea gigantifolia, C. sp. nov., C. rubriacuta, and C. moorei were datasets, but complete data for the remaining taxa. These are all discrepant in their phylogenetic relationships between cpDNA referred to as comb + cpDNA and comb + nrDNA, respectively and nrDNA. The placement of each of the species was constrained herein. to the topology of one of the partitions and then tested against the Double-stranded DNA was amplified via PCR, following the data of the other with no other changes made in the topology of methods of Smith et al. (1997). Sequences were obtained either the tree. For example, the nrDNA tree was used, except Columnea through the methods of Smith et al. (2004) or through Genewiz moorei was constrained to be part of the section Columnea clade, (South Plainfield, NJ), chromatograms were viewed and sequences where it was recovered from cpDNA data and then this topology edited and aligned by hand in PhyDE (Müller et al., 2005). was tested against the recovered nrDNA topology and SH (Shimodaira and Hasegawa, 1999) and AU (Shimodaira, 2002) tests 2.2. Test of incongruence were conducted. This was done for each of the four species sepa- rately and testing all topology/data combinations as well as con- The partition homogeneity test (PHT: Farris et al., 1994) was straining C. sp. nov., C. gigantifolia, and C. rubriacuta to be a single performed as implemented in PAUP⁄ v4.0 b10 (Swofford, 2002) clade, and all four species to the topology of the other dataset. with 10,000 bootstrap replicates (using a heuristic search, simple The SH tests were conducted in PAUP⁄ using full optimization addition, and no branch swapping). As an additional measure of and 1000 replicates. The AU tests were conducted in Consel with congruence among partitions, bootstrap analyses were performed the site likelihood values generated from PAUP⁄. on each partition separately to assess areas of conflict and to deter- The GSI (Cummings et al., 2008) was run on both the nrDNA and mine if any conflict was strongly supported (>70% support; cpDNA datasets separately as well as a concatenated dataset that Seelanan et al., 1997). Specific sequences that could be individually either included missing values for the four discrepantly placed spe- identified as incongruent with other partitions were identified as cies for nrDNA (comb + cpDNA) or cpDNA (comb + nrDNA). Colum- potentially from hybrid species. The PHT was also performed using nea gigantifolia, C. sp. nov., C. rubriacuta, C. moorei, C. citriflora, C. the same parameters on a dataset that excluded the putative dissimilis, C. strigosa, and C. minutiflora were each tested in terms hybrid taxa. of monophyly. Columnea citriflora, C. dissimilis, C. strigosa and C. minutiflora were tested because we had multiple accessions and 2.3. Phylogenetic analyses they serve as a means of evaluating the GSI results when a species is already expected to be monophyletic. An additional analysis was Phylogenetic trees were estimated using maximum parsimony run that constrained C. gigantifolia, C. rubriacuta, and C. sp. nov. to (MP), maximum likelihood (ML) and Bayesian inference (BI) for be inclusively monophyletic. We used the online version J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 231
Fig. 1. Representive photos of the putative hybrids, B. Columnea sp. nov., E. C. gigantifolia,H.C. rubriacuta, and K. C. moorei. Representative photos of clade A; A. C. picta, D. C. schimpfiii, and G. C. asteroloma. Representative photos of clade C; C. C. rubricalyx, F. C. herthae, and I. C. fimbricalyx. Additionally, C. minor (L), and C. scandens as a representative of clade B. (J) both as putative parents of C. moorei.
(Cummings et al., 2008; http://molecularevolution.org/soft- monophyletic) to 1 (monophyletic). Interpreting values between ware/phylogenetics/gsi/citation). We manually pruned the results these two is not always clear and we set an a priori value of of the Bayesian analyses to the final 500 sampled trees to reduce GSI > 0.80 to indicate monophyly. computational time. The analyses used all trees for 10,000 permu- To further discriminate between incomplete lineage sorting and tations. The analyses produce a GSI value that ranges from 0 (not hybridization we analyzed our data with JML (Joly, 2012). As input, 232 J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 the analysis requires a posterior distribution of species trees gener- 3.3. Splitstree ated from ⁄BEAST (Heled and Drummond, 2010). We generated species trees for nrDNA and cpDNA with the same species defini- The splitstree results place the putative hybrid taxa in agree- tions used for GSI for 10 million generations, selecting the ‘‘piece- ment with the cpDNA or nrDNA independent phylogenetic analy- wise constant” option for the population size model as ses (Figs. 2 and 3). There is a considerable amount of network recommended by Joly (2012). We made comparisons across spe- connecting many of the branches, including alternative arrange- cies trees (here actually gene trees because we analyzed each data- ments for the placement of the putative hybrid species. However, set separately) against the two data sets independently (i.e. cpDNA the branches leading to the clades that contain the hybrid taxa tree and simulated cpDNA data with observed cpDNA data as well are reasonably long (Fig. 3A and C). The splitstree of the combined as cpDNA tree and simulated cpDNA data with observed nrDNA dataset places the putative hybrid taxa in branches that are inter- data). The heredity scalar was set to 0.5 for the cpDNA simulated mediate to their placement in the cpDNA and nrDNA independent data and 2.0 for the nrDNA simulated data. The seqgencommand analyses (Fig. 3B). A total of 221 splits are present in the analysis used the model that was recommended from jmodeltest for the that contains all data, but only 168 are present when the putative observed cpDNA and nrDNA separately. All other factors used the hybrid taxa individuals are removed from the data. Although splits default option. Because so many comparisons are made, caution are still present, they are primarily near the center of network must be used to interpret significance of the results. We follow (where support for relationships is low, Fig. 2) and the branches the advice of Joly (2012) by evaluating only species-pair compar- leading to the different clades are reasonably long as they are in isons that are relevant to the discrepant taxa. Namely Columnea Fig. 3A and C. gigantifolia, C. rubriacuta, C. sp. nov. and C. moorei simulated pair- wise distances vs. the observed pairwise distances in clades where 3.4. SH/AU tests they are placed with their cpDNA (C. asteroloma, C. picta, C. eburnea, C. schimpfii, and C. densibracteata for C. gigantifolia, C. rubriacuta and The SH and AU tests gave almost equivalent results (Table 1). C. sp. nov.; C. arguta, C. bilbergiana, C. gloriosa, C. microphylla and C. When the topology resulting from one dataset was tested against scandens for C. moorei) or nrDNA (C. minutiflora, C. herthae, C. lucifer, the data from the other (e.g. cpDNA tree against the nrDNA data) C. pygmaea, C. rubricalyx, and C. fimbricalyx for C. gigantifolia, C. most placements were significantly different for both tests. The rubriacuta and C. sp. nov.; C. minor for C. moorei). exceptions were the placement of C. sp. nov., C. gigantifolia, and C. moorei for the SH tests. The AU tests were significant for all alter- 3. Results native topologies. The GSI results indicated that with nrDNA (including nrDNA+) 3.1. Incongruence test all individual species that were tested were monophyletic (GSI > 0.80) and significantly so with the exception of C. moorei The PHT indicated significant incongruence between the cpDNA (Table 2). In contrast, the cpDNA (including cpDNA+) only recov- and nrDNA datasets when all taxa were included (p < 0.001), but ered C. sp. nov., C. citriflora, C. dissimilis, C. strigosa, and C. minuti- not on the dataset that excluded the putative hybrid individuals flora as monophyletic (the latter four were not in question, but (p = 0.065). A comparison of MP bootstrap support between the including multiple individuals allowed us to test the monophyly trees derived from the separate datasets indicated that the place- of these species as well). Columnea gigantifolia and C. moorei were ment of C. rubriacuta, C. gigantifolia, C. sp. nov., and C. moorei were not monophyletic (GSI = 0.25, 0.268, respectively) and C. rubriacuta in conflict with at least some (BS > 70) support (Fig. 2). was below our accepted value of GSI > 0.80, but is closer (GSI = 0.77). The JML analyses did not indicate any significance when simu- 3.2. Phylogeny lated distances for a dataset were compared against the same observed data (e.g., simulated cpDNA distances against observed The results of the MP analysis of the cpDNA dataset resulted in cpDNA data). The simulated cpDNA distances were not signifi- 144 equally parsimonious trees of 415 steps each, CI = 0.65, cantly shorter than observed nrDNA data implying that the dis- RI = 0.86, RC = 0.71 (Fig. 2). Columnea rubriacuta, C. gigantifolia crepancies caused by cpDNA could be explained by incomplete and C. sp. nov. were recovered in clade A (Fig. 2) that includes all lineage sorting. In contrast simulated nrDNA data generated some of the sampled species from section Collandra. Columnea moorei distances that were significantly shorter than observed cpDNA dis- was recovered as part of clade B (Fig. 2) that includes all sampled tances (Table 3) implying nrDNA capture. Only comparisons of the species of section Columnea. Analyses of the dataset that included simulated nrDNA data and observed cpDNA placements are pre- all combined data, but missing nrDNA sequences for the four spe- sented as all other results were not significant. cies mentioned above was less resolved and with lower support for resolution in the MP, ML and BI trees (results not shown). The nrDNA dataset yielded 196 equally parsimonious trees of 4. Discussion 614 steps each, CI = 0.52, RI = 0.81, RC = 0.49 (Fig. 2). Columnea rubriacuta, C. gigantifolia, and C. sp. nov. were recovered as part Significant differences between the topologies of the cpDNA of clade C (Fig. 2). Columnea moorei was recovered as sister to C. and nrDNA datasets were detected using the partition homogene- minor (Fig. 2). Analyses that included all combined data, but miss- ity test and were found to be the result of hard incongruence ing sequences for cpDNA for the four species mentioned above was (Seelanan et al., 1997; Fig. 2). Specifically, the incongruence could approximately the same as the dataset that included only nrDNA in be traced to the inclusion of four species; Columnea gigantifolia, C. terms of topology, resolution, and support for resolution (results sp. nov., C. rubriacuta, and C. moorei. When these species were not shown). Convergence was achieved in all BI results based on excluded from the analyses no incongruence was detected as ESS values all over 2000 at a minimum as viewed in Tracer and determined by bootstrap support over 70% for alternative place- comparison of separate runs in AWTY indicate that the analyses ment of individuals between the datasets (results not shown). Fur- were approximating the same target distribution (results not ther evidence for the incongruence is supported by SH and AU tests shown). The ML and majority rule BI trees for all datasets were that compared the placement of individual and groups of species congruent with the MP tree, but with greater resolution (Fig. 2). from one dataset to the data of the other. In nearly all comparisons J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 233
Fig. 2. Majority rule consensus tree from the Bayesian inference of A. nrDNA data. B. cpDNA. The topology of these trees are congruent with the bootstrap consensus trees from maximum parsimony and maximum likelihood and therefore both maximum parsimony bootstrap/maximum likelihood bootstrap/Bayesian inference posterior probability values are presented. Missing values (À) indicate clades not supported by bootstrap over 50. Thick lines indicate support of >75/>75/>95, double thick lines indicate support >90/> 90/100. Species names in blue are where the putative hybrids fall with nrDNA, names in red are where the same species fall in cpDNA and putative hybrids are in purple font. 234 J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240
Fig. 3. Splitstree representation using the neighbornet optimization of A. cpDNA, B. combined cpDNA and nrDNA, and C. nrDNA. Taxon names have been removed for simplicity, but are available in supplementary Figs. 1–3. Violet shapes are the putative hybrid species discussed in the text. Blue-purple shapes are species of clades A and B (Fig. 2) and peach shapes are species of clade C (Fig. 2)orC. minor.
the tests were significant (Table 1) indicating conflict between the Table 1 Results of SH and AU tests for the placement of species from either of the datasets data sets. analyzed independently against the data from the other dataset. The conflict is further demonstrated with the splitstrees using neighbornet. The individual datasets place the putative four hybrid Taxon cpDNA position and nrDNA position and nrDNA data; SH/AU cpDNA data; SH/AU species with the species in the clade that they fell out with in phy- logenetic analyses (Figs. 2 and 3). There are many connections in C. sp. nov. 0.373/<0.001 <0.187/<0.001 the splitstrees for these two datasets, especially toward the center C. gigantifolia 0.052/0.05 <0.219/0.05 C. rubriacuta <0.001/<0.001 <0.003/0.023 (Fig. 3A and C) or what would be the ‘‘backbone” of the phyloge- C. sp. nov., C. gigantifolia, <0.001/<0.001 <0.005/0.006 netic analyses where support for relationships also tends to be and C. rubriacuta low (Fig. 2). However, the branches leading to the placement of C. moorei 0.262/<0.001 0.071/<0.001 the putative hybrids are reasonably long, at least longer than many All of the above <0.001/<0.001 <0.001/<0.001 of the other internal branches indicating that the discrepancies J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 235
Table 2 GSI results to test the monophyly of species or clades that were not recovered as monophyletic or with poor support in the phylogenetic analyses. Datasets were either the nrDNA or cpDNA data alone or nrDNA+ that included the combined data, but missing cpDNA for the putative hybrids and cpDNA+ that include the combined data, but missing nrDNA for the putative hybrids. Values reported are the GSI which ranges from 0 to 1 and the statistical p value.
Taxon nrDNA data: GSI/p-value cpDNA data: GSI/p-value nrDNA+ GSI/p value cpDNA+ GSI/p value C. sp. nov. 1.00/0.0001 0.82/0.0003 1.00/0.001 0.82/0.0002 C. citriflora 0.79/0.0001 1.00/0.0001 0.78/0.0001 1.00/0.0001 C. dissimilis 1.0/0.0017 0.90/0.008 1.0/0.001 0.89/0.006 C. gigantifolia 1.00/0.0022 0.25/0.03 1.00/0.002 0.25/0.04 C. moorei 0.44/0.013 0.268/0.03 0.45/0.02 0.27/0.04 C. rubriacuta 0.88/0.0001 0.77/0.0001 0.912/0.001 0.77/0.001 C. strigosa 0.90/0.0036 0.98/0.002 0.99/0.009 0.98/0.006 C. minutiflora 1.00/0.0023 1.00/0.003 1.00/0.006 1.00/0.003 C. sp. nov., C. gigantifolia, and C. rubriacuta 0.77/0.0001 0.61/0.0001 0.76/0.0001 0.613/0.0001
Table 3 nrDNA and they were able to demonstrate that in species where Results of the JML test showing the probabilities of obtaining simulated nrDNA phylogenetic discrepancy existed between cpDNA and nrDNA that distances compared to observed cpDNA distances without hybridization. Multiple the cpDNA recovered monophyletic species whereas the nrDNA values were obtained for some comparisons as a result of sampling more than one individual per species. Only the putative hybrid species, in comparison to the species did not, implying that the cpDNA had been captured in an ancestral in the clades where they fall with cpDNA were compared. Only distances where the event. Our results indicate the reverse for at least C. gigantifolia, C. probability is less than 0.1 are shown, all other results were insignificant including all rubriacuta, and C. moorei where the nrDNA GSI results recovered comparisons made of simulated cpDNA distances compared to nrDNA distance. Other monophyletic species, but the cpDNA did not. Given that the rate distances, for example any putative hybrid species and species not in any of the clades where the putative hybrids fell, were not examined. of coalescence for cpDNA should be much faster than nrDNA, the data imply that nrDNA must have been transferred to these species Species pairs and has coalesced to an ancestral copy within each of the species. C. sp. nov.-C. asteroloma 0.0001 This has not occured for the cpDNA, which implies that coalescence C. sp. nov.-C. densibracteata 0.0001 has not yet occured within each of the species. If there were no C. sp. nov.-C. eburnea 0.0001–0.0003 C. sp. nov.-C. schimpfii 0.0002–0.0026 cross-species transfer of either material, the reverse might have C. sp. nov.-C. picta 0.01–0.039 occured by chance. In contrast, the species where the GSI was C. gigantfolia-C. asteroloma 0.0001 scored, but were recovered as monophyletic in all analyses have C. gigantfolia-C. densibracteata 0.0001 a greater GSI for cpDNA than nrDNA as would be expected when C. gigantfolia-C. eburnea 0.0001 comparing a haploid genome to a diploid one (Table 2). C. gigantfolia-C. schimpfii 0.0005 C. gigantfolia-C. picta 0.006 Coalescence of nrDNA and not cpDNA implies a bottleneck in C. rubriacuta-C. asteroloma 0.0001–0.046 the ancestor to these individuals in terms of nrDNA, but not C. rubriacuta-C. densibracteata 0.0001–0.025 cpDNA. An explanation could be interspecific hybridization where C. rubriacuta-C. eburnea 0.0002–0.09 the F1 hybrid either self-fertilized, or continued to backcross with C. rubriacuta-C. schimpfii 0.0018–0.06 C. rubriacuta-C. picta 0.017–0.067 other members of the maternal parent, or maternal parent clade, C. moorei-C. gloriosa 0.0002 and by chance the paternal copy of the nrDNA became fixed. This C. moorei-C. microphylla 0.0002–0.04 implies nrDNA capture. These data are in agreement with the mor- C. moorei-C. scandens 0.0002–0.058 phology for C. gigantifolia, C. rubriacuta, and C. sp. nov., all of which would be placed in clade A (Figs. 1 and 2) that corresponds to sec- tion Collandra based on the sampling here using morphology alone. between the two datasets is not the result of poor or insufficient These species are distinctive by having dorsoventral shoots, data in one or both datasets (Morrison, 2009). The splitstree based strongly anisophyllous oblanceolate leaves that often are colored on the combined data places the putative hybrid taxa in a network red or purple on the abaxial surface, either entirely or with spots that is intermediate to their placements in both the cpDNA and or marginal coloration, calyces that are green or yellow in color, nrDNA (Fig. 3B). and large overlapping bracts subtending the flowers (Kvist and Interpreting phylogenetic incongruencies between datasets can Skog, 1993). Moreover, these species have a dispersed spatial dis- be challenging as they may result from paralogy, incomplete lin- tribution, produce a few flowers per plant throughout the year, eage sorting, horizontal gene transfer, or hybridization (Buckley and are pollinated exclusively by hermit hummingbirds Also, the et al., 2006; Kubatko and Degnan, 2007; Holland et al., 2008; plants tend to have and autoincompatible reproductive system Maureira-Butler et al., 2008; Joly et al., 2009; Pirie et al., 2009; (Marín-Gómez, 2014; Marín-Gómez and Amaya-Márquez, 2015). Polihronakis, 2009; Willyard et al., 2009; Pelser et al., 2010; de They do not fit with the morphology of species in clade C (Figs. 1 Viliers et al., 2013; Yu et al., 2011, 2013). Paralogy can be elimi- and 2) that although anisophyllous, are elliptic and with crenate nated for the sequences analyzed here. Chloroplast DNA is haploid margins, calyces are often red-orange in color and bracts subtend- and although concerted evolution has been documented to be ing the flowers are small or lacking. incomplete or lacking in some nrDNA for other members of Gesne- The JML test was developed as a means of detecting hybridiza- riaceae (Denduangboripant and Cronk, 2000), all of the ITS and ETS tion or incomplete lineage sorting. The program simulates a set of sequences generated for this study were done using direct DNA sequences based on trees generated by real data and then sequencing and results were at least as clean as the haploid cpDNA compares the distances between species pairs based on the simu- regions. lated vs. observed data. Species pairs that are less distant with the De Viliers et al. (2013) investigated potential interspecific simulated data compared to the observed data imply that hybridization in Streptocarpus species that also demonstrated dis- hybridization has occurred. In this study, the simulated cpDNA dis- crepancies between datasets. They used the GSI as an indicator that tances were not significantly shorter than observed nrDNA data coalescence had occurred in the lineage for either the nrDNA or implying that the discrepancies caused by cpDNA could be cpDNA data. Haploid cpDNA will coalesce faster than diploid explained by incomplete lineage sorting. In contrast simulated 236 J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 nrDNA data generated some distances that were significantly Columnea clade as maternal parents, or that F2 or backcross shorter than observed cpDNA distances (Table 3) implying nrDNA hybrids were generated that resulted in the cultivated material capture. These results are consistent with the GSI results in that a of C. moorei that we have now. Therefore, we cannot assume mul- lack of coalescence for the cpDNA would result in simulated tiple maternal parents for C. moorei. sequences that would be unlikely to be shorter than nrDNA dis- Determining the paternal parent of C. moorei is also a challenge. tances that resulted from a more recent coalescence. However, Both accessions of C. moorei are with C. minor in the nrDNA phylo- the JML test was also able to detect shorter distances for C. sp. genetic analyses (Fig. 2). Although C. minor is a widespread and nov. that GSI recovered as monophyletic for both datasets (Tables common species in Colombia and Ecuador, it is currently not 2 and 3). known from Panama. It is also possible that the closest relative to the nrDNA of C. moorei remains unsampled in our analyses. 4.1. Sympatry of parental species Columnea minor is recovered as sister to a clade of Jamaican species in analyses that exclude C. moorei (Smith et al., 2013), but is placed It is imperative that if hybridization is occurring, that the paren- within section Angustiflorae here (Fig. 2) when C. moorei is tal species are sympatric, or at least were sympatric at the time of included. Analyses that used greater sampling of section Angustiflo- hybridization. Our analyses do not have the resolving power to link rae were able to exclude C. moorei as part of that clade (Schulte any of the hybrids to parental species in either group (Fig. 2). Like- et al., 2014) implying that the placement, and possibly the close wise, our sampling was not conducted as a means of determining relationship of C. moorei to C. minor with nrDNA is an artefact of the exact parental species. In fact, section Collandra, where C. taxon sampling. gigantifolia, C. rubriacuta, and C. sp. nov. are placed with cpDNA is It does not appear that C. moorei is simply an F1. If it were a the largest section in the genus in terms of number of species stable F1 hybrid, then it would be expected to have nrDNA from and may include over 80 species, whereas we have only sampled both parents. All sequencing of nrDNA for this species was done five here (Fig. 2). However, we do know that where C. gigantifolia, directly and all sequences produced clean reads without multiple C. rubriacuta, and C. sp. nov. are found in their present day, that peaks. This implies that either after the initial hybridization event other species from both of the parental clades are also found. In self-fertilization resulted in the nrDNA of the C. minor lineage many cases this can include multiple species from both clades becoming fixed and the nrDNA from the maternal lineage was lost. (pers. obs.). Alternatively, it is possible that outcrossing with other members of Columnea moorei presents a different challenge in terms of iden- clade B continued with the ancestor to the modern C. moorei with tifying putative parental species. This species has a narrow range, backcrossing to the hybrid or selfing such that only the nrDNA of found only on Cerro Jefe in Panama. Species of section Columnea, the C. minor lineage was retained. This latter situation is likely where C. moorei is placed with cpDNA are diverse in Panama, given that C. moorei is not recovered as monophyletic in any of and Central America in general (Skog, 1978). We were able to the analyses and may reflect a stable morphological species that obtain recently collected wild material of this species from Cerro is the result of several genetic lineages. Jefe, along with all other species of Columnea that are currently known from that region including C. arguta and C. billbergiana from 4.2. Hybrid lineage for C. gigantifolia, C. rubriacuta, and C. sp. nov. Cerro Jefe (Appendix A). The phylogenetic analyses of cpDNA place the recently collected wild material of C. moorei as sister to C. bill- The most parsimonious explanation for the hybrid origin of C. bergiana (Fig. 2), implying that C. billbergiana may be the maternal gigantifolia, C. rubriacuta, and C. sp. nov. is an ancestral hybridiza- parent. However, the cultivated specimen of C. moorei was not sis- tion event with the pollen donor being a species of clade C (Figs. 1 ter to C. billbergiana and is instead sister to a clade of C. gloriosa/C. and 2) and the maternal lineage a member of clade A (Figs. 1 and microphylla (Fig. 2). These data, and the low GSI values recovered 2). This F1 hybrid likely continued to backcross either with other for both cpDNA and nrDNA (Table 2) for C. moorei suggest that individuals of its maternal species, or other members of clade A hybridization may still be active with this species. However, the and by chance the nrDNA from its paternal parent became fixed, cultivated specimen of C. moorei may have also undergone more whereas variability in the cpDNA genome persisted. This ancestral recent hybridization. This species has seldom been collected in species then gave rise to the three species sampled here, and per- the wild (five are cited in Skog, 1978). One of these is presumably haps others. Further support for this stems from the GSI results, the voucher of the wild plant for the material in cultivation that which indicate that all three species as a clade approach our was used here. This individual was used as a parent in a series of threshold of 0.8 for being monophyletic with nrDNA (GSI = 0.77, crosses with other Columnea species in cultivation at Cornell Table 2), but are less likely to be a monophyletic group with cpDNA University in the 1960s. It is possible that the material in cultiva- (GSI = 0.61, Table 2). None of the three species we have recovered tion represents the results of one of these crosses or a potential as being of hybrid origin were previously suspected to be hybrids F2 or backcrossed hybrid. A series of vouchers at Cornell University and it is possible that this lineage includes other members of sec- document the F1 hybrids that were generated with C. pilosissima tion Collandra that have yet to be sampled with molecular meth- (currently considered synonymous with C. hirta), C. percrassa (cur- ods. However, we cannot preclude the possibility of three rently considered synonymous with C. billbergiana), C. nicaraguen- independent hybridization events for the origin of these species sis Oerst., C. linearis Oerst., C. illepida H.E. Moore, C. gloriosa, C. as the GSI results are not exceptionally strong in recognizing this glabra Oerst., and C. allenii C.V. Morton. All but C. illepida are con- clade as monophyletic. sidered members of section Columnea, represented in our analyses by the species of clade B, where C. moorei is placed with the cpDNA Acknowledgments data. In many cases the vouchers clearly state that C. moorei was the female parent, and in all cases C. moorei is listed first. Assuming The authors would like to thank the following for their generos- that the annotations are consistent then the cpDNA of these ity in sharing plant material: Jerry Harrison, Larry Skog, Nancy and hybrids would have reflected the cpDNA of C. moorei and not the Jerry Kast, Bob Stewart, and Bill Price. Special thanks go to Simon pollen donor species. However, we cannot rule out that other Joly for assistance with files when the authors failed to get the crosses were made that we no longer have documentation of formatting correct for JML. Funding for this project is from NSF – (potentially also including C. minor which C. moorei is sister to with United States, grant DEB0949270 to JFS and JLC. MAM thanks the the nrDNA data, Fig. 2), that included members of the section National University for time to do research. To the Organización Table A1 Species, voucher specimens, and GenBank accession numbers for all accessions included in phylogenetic analyses. NA indicates that sequences were not generated for this region for this particular accession. ex indicates a sequence that was excluded from the analysis due to incongruence with other sequences. Accesion numbers in bold were newly generated for this study (to be added [TBA] when final acceptance has been indicated).
Taxon Voucher Herbarium Collection Locality rpl32- trnQ-rps16 rps16 trnS-G trnH-psbA ITS ETS trnL UAG intron intron spacer spacer spacer Columnea albovinosa (M. Freiberg) J. L. Clark 7386 US Ecuador KX912467 KX912524 KX912581 NA KX912411 KX912354 KX912276 J.L. Clrk & L. E. Skog C. angustata (Wiehler) L. E. Skog J. L. Clark 8627 UNA & US Panama KF005816 KF006034 KF005925 KF006137 NA KF005727 KP260806 C. anisophylla DC. J. F. Smith 10773 COL Colombia KX912468 KX912525 KX912582 KX912635 KX912412 KX912355 KX912277 C. arguta C. V. Morton J. & L. Harrison 685 SRP Panama KX912469 KX912526 KX912583 KX912636 KX912413 KX912356 KX912278 C. asteroloma (Wiehler) L.E. Skog J. L. Clark 7950 US Ecuador KX912470 KX912527 KX912584 KX912637 KX912414 KX912357 KX912279 C. billbergiana Beurl. J. & L. Harrison 683 SRP Panama KX912471 KX912528 KX912585 KX912638 KX912415 KX912358 KX912280 C. brenneri (Wiehler) B. D. Morley J. F. Smith 3385 SRP Ecuador KF005823 KF006040 KF005932 KF006144 KF005649 KF005734 KX912281 C. brevipila Urb. J. F. Smith 10058 SRP Cultivated, Jamaica KF005825 KF006042 KF005934 KF006146 KF005651 KF005736 KX912282 C. byrsina (Wiehler) L. P. Kvist & L. E. Skog J. F. Smith 3408 SRP Ecuador KF005826 KF006043 KF005935 JQ953714 KF005652 KF005737 KP260812 C. calotricha Donn. Sm. J. F. Smith et al. 4117 SRP French Guiana KF005828 KF006045 KF005937 KF006149 KF005654 KF005739 KX912283 228–240 (2017) 106 Evolution and Phylogenetics Molecular / al. et Smith J.F. C. ciliata (Wiehler) L.P. Kvist & L.E. Skog J. F. Smith 8604 SRP Cultivated, Ecuador KX912472 KX912529 KX912586 KX912639 KX912416 KX912359 KX912284 C. citriflora L. E. Skog J. L. Clark 10053 UNA & US Cultivated, Panama KF005830 KF006047 KF005939 KF006151 KF005655 KF005741 KX912286 C. citriflora L. E. Skog J. L. Clark 10453 US Cultivated, Panama KX912474 KX912531 KX912588 KX912641 KX912418 KX912361 KX912287 C. citriflora L. E. Skog J. & L. Harrison 682 SRP Panama KX912473 KX912530 KX912587 KX912640 KX912417 KX912360 KX912285 C. cruenta B.D. Morley J. & L. Harrison 686 SRP Panama KX912475 KX912532 KX912589 KX912642 KX912419 KX912362 KX912288 C. dielsii Mansf. J. F. Smith 1989 WIS Ecuador KF005836 KF006053 KF005945 KF006157 KF005661 KF005747 KP260822 C. dissimilis C. V. Morton J. & L. Harrison 684 SRP Panama KX912476 KX912533 KX912590 KX912643 KX912420 KX912363 KX912288 C. dissimilis C. V. Morton J. L. Clark 7500 UNA Panama KX912477 KX912534 KX912591 KX912644 KX912421 KX912364 KX912290 C. eburnea (Wiehler) L.P. Kvist & L.E. Skog J. L. Clark 6353 US Ecuador KF005840 KF006057 KF005949 KF006160 KF005665 KF005750 KX912291 C. elongatifolia L.P. Kvist & L.E. Skog J. L. Clark 10015 US Ecuador KF005841 KF006058 KF005950 KF006161 KF005666 KF005751 KX912292 C. fimbricalyx L.P. Kvist & L.E. Skog J. L. Clark 10429 US Ecuador KX912478 KX912535 KX912592 KX912645 KX912422 KX912365 KX912293 C. flexiflora L.P. Kvist & L.E. Skog O.J. H. L. Marín-Gómez Clark & L. Jost 289 6968 COLUS Colombia Ecuador KX912479 KF005846 KX912536 KF006063 KX912593 KF005956 KX912646 KF006167 KX912423 KF005671 KX912366 KF005755 KX912295 KX912294 C.C. gigantifolia gigantifoliaL.L. P. P. Kvist Kvist & & L. L. E. E. Skog Skog O. H. Marín-Gómez 296 COL Colombia KX912480 KX912537 KX912594 KX912647 KX912424 KX912367 KX912296 C. gloriosa Sprague J. L. Clark et al. 9921 US Ecuador KF005848 KF006065 KF005958 KF006169 KF005673 KF005757 KX912297 C. herthae Mansf. J. L. Clark 7113 US Ecuador KF005853 KF006069 KF005963 KF006173 KF005677 KF005761 KX912299 C. hypocyrtantha (Wiehler) J. L. Clark & E. Rodriguez US Bolivia KF005854 KF006071 KF005964 KF006174 KF005679 KF005762 KX912300 J.F. Smith & L.E. Skog 6741 C. isernii Cuatrec. J. L. Clark et al. 6253 US Ecuador KF005857 KF006074 KF005967 KF006177 DQ211220 AF543247 KX912301 C. kucyniakii Raymond T. Croat 94640 MO Ecuador KX912481 KX912538 KX912595 KX912648 KX912425 KX912368 KX912302 C. lehmannii Mansf. J. L. Clark 13267 US Colombia KX912482 KX912539 NA KX912649 KX912426 KX912369 KX912303 C. lophophora Mansf. J. L. Clark et al. 7888 US Ecuador KF005860 KF006076 KF005969 KF006179 KF005684 KF005767 KP260825 C. lucifer J.L. Clark J. L. Clark 11100 US Ecuador KX912483 KX912540 KX912596 KX912650 KX912427 KX912370 KX912304 C. medicinalis (Wiehler) L.E. Skog & L.P. Kvist T. Croat 94600 MO Ecuador KX912484 KX912541 KX912597 KX912651 KX912428 KX912371 KX912305 C. microphylla Klotsch & Hanst. J. L. Clark 6261 UNA & US Cultivated KF005863 KF006080 KF005973 KF006182 KF005687 KF005771 KP260827 (ex) C. minor (Hook.) Hanst. T. Croat 94778 MO Ecuador KF005866 KF006084 KF005975 KF006185 KF005690 KF005774 KP260828 (ex) C. minutiflora L. P. Kvist & L. E. Skog J. L. Clark et al. 7092 UNA & US Ecuador KF005868 KF006086 KF005977 KF006187 KF005692 KF005776 KX912306 C. minutiflora L. P. Kvist & L. E. Skog O. H. Marín-Gómez 318 COL Colombia KX912485 KX912542 KX912598 KX912652 KX912429 KX912372 KX912307 C. moesta Poepp. J. F. Smith 1829 WIS Bolivia KF005870 KF006084 KF005979 KF006189 KF005694 KF005778 KP260830 C. moorei C. V. Morton J. & L. Harrison 687 SRP Panama KX912486 KX912543 KX912599 KX912653 KX912430 KX912373 KX912308 C. moorei C. V. Morton J. L. Clark 11307 UNA & US Cultivated, originally KX912487 KX912544 KX912600 KX912654 KX912431 KX912374 KX912309 from Panama C. oblongifolia Rusby J. F. Smith 1725 WIS Bolivia KF005874 KF006092 KF005983 KF006193 KF005697 KF005781 KX912310 C. orientandina Mansf. J. L. Clark et al. 9885 UNA Ecuador KF005876 KF006094 KF005985 KF006195 KF005699 KF005783 KP260833 C. paramicola (Wiehler) L.P. Kvist & Not vouchered na Cultivated at Smithsonian as KF005878 KF006095 KF005987 JQ954064 DQ211224 DQ211113 KX912311 L.E. Skog USBG94529, Ecuador C. picta H. Karst. T. Croat 94956 MO Ecuador KF005879 KF006096 KF005988 KF006197 KF005701 KF005785 KP260837
C. pulchra C.V.J.L. Clark Morton & J.F. Smith J.J. L. L. Clark Clark 11180 6265 UNAUS & US Ecuador Cultivated KX912488 KF005880 KX912545 NA KX912601 KF005990 KX912655 KF006198 KX912432 DQ211225 KX912375 KF005786 KX912313 KX912312 237 C. pygmaea (continued on next page) 238 Table A1 (continued)
Taxon Voucher Herbarium Collection Locality rpl32- trnQ-rps16 rps16 trnS-G trnH-psbA ITS ETS trnL UAG intron intron spacer spacer spacer C. repens (Hook.) Hanst. J. F. Smith 8605 SRP Cultivated, Jamaica KF005884 KF006100 KF005993 KF006201 KF005705 KF005790 KX912314 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 4510 UNA & US Ecuador KX912501 KX912558 KX912612 KX912664 KX912444 KX912388 KX912327 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 4975 QCNE & US Ecuador KX912502 KX912559 KX912613 KX912665 KX912445 KX912389 KX912328 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 7493 UNA & US Ecuador KX912503 KX912560 KX912614 KX912666 KX912446 KX912390 KX912329 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 8807 US Ecuador KX912504 KX912561 KX912615 KX912667 KX912447 KX912391 KX912330 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 9622 US Ecuador KX912505 KX912562 KX912616 KX912668 KX912448 KX912392 KX912331 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 10122 US Ecuador KX912491 KX912548 KX912604 KX912658 KX912435 KX912378 KX912317 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 10386 US Ecuador KX912492 KX912549 KX912605 NA KX912436 KX912379 KX912318 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 10387 US Ecuador KX912493 KX912550 KX912606 NA KX912437 KX912380 KX912319 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 10416 US Ecuador KX912494 KX912551 NA KX912659 KX912438 KX912381 KX912320 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 10445 US Ecuador KX912495 KX912552 NA KX912660 KX912439 KX912382 KX912321
C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 11063 US Ecuador KX912496 KX912553 KX912607 NA KX912440 KX912383 KX912322 228–240 (2017) 106 Evolution and Phylogenetics Molecular / al. et Smith J.F. C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 13187 US Colombia KX912497 KX912554 KX912608 KX912661 KX912441 KX912384 KX912323 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 13367 US Colombia KX912498 KX912555 KX912609 KX912662 KX912442 KX912385 KX912324 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 13401 US Colombia KX912500 KX912557 KX912611 KX912663 KX912443 KX912387 KX912326 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. L. Clark 13388 US Colombia KX912499 KX912556 KX912610 NA NA KX912386 KX912325 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog T. C. Croat 95134 MO Ecuador KX912515 KX912572 KX912626 KX912678 KX912458 KX912398 KX912341 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 259 COL Colombia KX912506 KX912563 KX912617 KX912669 KX912449 KX912403 KX912332 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 264 COL Colombia KX912507 KX912564 KX912618 KX912670 KX912450 KX912393 KX912333 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 273 COL Colombia KX912508 KX912565 KX912619 KX912671 KX912451 KX912394 KX912334 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 278 COL Colombia KX912509 KX912566 KX912620 KX912672 KX912452 KX912395 KX912335 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 303 COL Colombia KX912510 KX912567 KX912621 KX912673 KX912453 KX912396 KX912336 COL Colombia KX912511 KX912568 KX912622 KX912674 KX912454 KX912404 KX912337 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 308 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 311 COL Colombia KX912512 KX912569 KX912623 KX912675 KX912455 KX912405 KX912338 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 316 COL Colombia KX912513 KX912570 KX912624 KX912676 KX912456 KX912406 KX912339 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog O. H. Marín-Gómez 319 COL Colombia KX912514 KX912571 KX912625 KX912677 KX912457 KX912397 KX912340 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. F. Smith et al. 10833 COL Colombia KX912489 KX912546 KX912602 KX912656 KX912433 KX912376 KX912315 C. rubriacuta (H. Wiehler) L. P. Kvist & L. E. Skog J. F. Smith et al. 10871 COL Colombia KX912490 KX912547 KX912603 KX912657 KX912434 KX912377 KX912316 C. rubricalyx L. P. Kvist & L. E. Skog J. L. Clark et al. 11034 UNA Ecuador KF005887 KF006103 KF005997 KF006204 KF005707 KF005792 KX912342 C. scandens L. J. L. Clark & S. G. Clark US Martinique KF005890 KF006106 KF005999 KF006207 KF005711 KF005795 KX912343 6541 C. schimpffii Mansf. J. F. Smith 8605 SRP Cultivated, Ecuador KF005892 KF006109 KF006001 KF006209 KF005713 KF005797 KX912344 C. spathulata Mansf. J. F. Smith 1853 WIS Ecuador KF005893 KF006110 KF006003 KF006211 KF005715 KF005798 KP260844 C. sp. J. L. Clark 7525 US Ecuador KX912520 KX912577 KX912631 NA KX912463 KX912402 KX912298 C. sp. Not vouchered MTJB Cultivated KX912516 KX912573 KX912627 KX912679 KX912459 KX912399 KX912345 C. sp. nov. J. F. Smith et al. 10775 COL Colombia KX912517 KX912574 KX912628 KX912680 KX912460 KX912400 KX912348 C. sp. nov. J. F. Smith et al. 10854 COL Colombia KX912518 KX912575 KX912629 KX912681 KX912461 KX912401 KX912346 C. sp. nov. J. F. Smith et al. 10891 COL Colombia KX912519 KX912576 KX912630 KX912682 KX912462 KX912407 KX912347 C. strigosa Benth. J. L. Clark 9069 US Ecuador KX912522 KX912579 KX912633 KX912684 KX912465 KX912409 KX912350 C. strigosa Benth. E. J. Tepe 2353 SRP Ecuador KX912521 KX912578 KX912632 KX912683 KX912464 KX912408 KX912349 C. tenella L.P. Kvist & L.E. Skog T. Croat 95108 MO Ecuador KX912523 KX912580 KX912634 KX912685 KX912466 KX912410 KX912351 C. tenensis (Wiehler) B.D. Morley J. F. Smith 3374 Ecuador KF005898 KF006115 KF006008 KF006216 KF005720 KF005804 KX912352 C. trollii Mansf. J. F. Smith 1723 WIS Bolivia KF005899 KF006117 KF006010 KF006218 KF005722 KF005805 KX912353 Glossoloma anomalum J. L. Clark J. F. Smith 3418 SRP Ecuador KF005912 KF006128 KF006021 KF006224 NA AF543225 KP260850 Glossoloma grandicalyx J. F. Smith 3417 SRP Ecuador KF005913 KF006129 KF006024 JQ953708 DQ211205 AF543218 KP260851 (J. L. Clark & L. E. Skog) J. L. Clark Glossoloma martinianum J. L. Clark 6101 US Ecuador KF005914 KF006130 KF006022 JQ953709 DQ211209 AF543228 KP260852 /ce:italic> (J. F. Smith) J. L. Clark Glossoloma panamense L. E. Skog et al. 7641 US Cultivated KF005915 KF006131 KF006023 JQ953710 DQ211202 DQ211102 KP260853 (C. V. Morton) J. L. Clark J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240 239
Ambiental Comunitaria SERRANIAGUA, César Franco Laverde, Doyle, J.J., 1992. Gene trees and species trees: molecular systematics as one- Johnier Arango Bermúdez, Milton Pineda Duque, Mauricio Florez character taxonomy. Syst. Bot. 17, 144–163. Farris, J.S., 1989. The retention index and the rescaled consistency index. Cladistics Pai and Reserva Natural Rio Ñambí for the logistic support pro- 5, 417–419. vided to the 2013 expedition. OHMG was supported by a student Farris, J.S., Källersjö, M., Kluge, A.G., Bult, C., 1994. Testing significance of grant of the Facultad de Ciencias de la Universidad Nacional de incongruence. Cladistics 10, 315–319. Feliner, G.N., Rosselló, J.A., 2007. Better the devil you know? Guidelines for Colombia, the Nellie D. Sleeth Scholarship (The Gesneriad Society insightful utilization of nrDNA ITS in species-level evolutionary studies in INC), and the programa de incentivos para la investigación Thomas plants. Mol. Phylogenet. Evol. 44, 911–919. van der Hammen (Jardín Botánico José Celestino Mutis). We thanks Felsenstein, J., 1985. Phylogenies and the comparative method. Am. Nat. 125, 1–15. to Herbario Nacional Colombiano (COL). Friar, E.A., Prince, L.M., Cruse-Sanders, J.M., McGlaughlin, M.E., Butterworth, C.A., Baldwin, B.G., 2008. Hybrid origin and genomic mosaicism of Dubautia scabra (Hawaiian silversword alliance; Asteraceae, Madiinae). Syst. Bot. 33, 589–597. Appendix A Givnish, T.J., 2010. Ecology of plant speciation. Taxon 59, 1326–1366. Givnish, T.J., 2015. Adaptive radiation versus ‘radiation’ and ‘explosive diversification’: why conceptual distinctions are fundamental to See Table A1. understanding evolution. New Phytol. http://dx.doi.org/10.1111/nph.13482. Glor, R.W., Losos, J.B., Larson, A., 2005. Out of Cuba: overwater dispersal and speciation among lizards in the Anolis carolinensis subgroup. Mol. Ecol. 14, Appendix B. Supplementary material 2419–2432. Hamilton, M.B., 1999. Four primer pairs for the amplification of chloroplast Supplementary data associated with this article can be found, in intergenic regions with intraspecific variation. Mol. Ecol. 8, 521–523. Hedrick, P.W., 2007. Sex differences in mutation, recombination, selection, gene the online version, at http://dx.doi.org/10.1016/j.ympev.2016.10. flow and genetic drift. Evolution 61, 2750–2771. 001. Heled, J., Drummond, A.J., 2010. Bayesian inference of species trees from multilocus data. Mol. Phylogenet. Evol. 27, 570–580. Holder, M.T., Anderson, J.A., Holloway, A.K., 2001. Difficulties in detecting References hybridization. Syst. Biol. 50, 978–982. Holland, B.R., Benthin, S., Lockhart, P.J., Moulton, V., Huber, K.T., 2008. Using supernetworks to distinguish hybridization from lineage-sorting. BMC Evol. Abbott, R., Albach, D., Ansell, S., Arntzen, J.W., Baird, S.J.E., Bierne, N., Boughman, J., Biol. 8, 202. Brelsford, A., Buerkle, C.A., Buggs, R., Butlin, R.K., Dieckmann, U., Eroukhmanoff, Howarth, D.G., Baum, D.A., 2005. Genealogical evidence of homoploid hybrid F., Grill, A., Cahan, S.H., Hermansen, J.S., Hewitt, G., Hudson, A.G., Jiggins, C., speciation in an adaptive radiation of Scaevola (Goodeniaceae) in the Hawaiian Jones, J., Keller, B., Marczewski, T., Mallet, J., Martinez-Rodriguez, P., Möst, M., islands. Evolution 5, 948–961. Mullen, S., Nichols, R., Nolte, A.W., Parisod, C., Pfennig, K., Rice, A.M., Ritchie, M. Huson, D.H., 1998. SplitsTree: analyzing and visualizing evolutionary data. G., Seifert, B., Smadja, C.M., Stelkens, R., Szymura, J.M., Väinölä, R., Wolf, J.B.W., Bioinformatics 14, 68–73. Zinner, D., 2013. Hybridization and speciation. J. Evol. Biol. 26, 229–246. Jabaily, R.S., Sytsma, K.J., 2010. Phylogenetics of Puya (Bromeliaceae): placement, Alice, L.A., Eriksson, T., Eriksen, B., Campbell, C.S., 2001. Hybridization and gene flow major lineages, and evolution of Chilean species. Am. J. Bot. 97, 337–356. between distantly related species of Rubus (Rosaceae): evidence from nuclear Joly, S., 2012. JML: testing hybridization from species trees. Mol. Ecol. Resour. 12, ribosomal internal transcribed spacer region sequences. Syst. Bot. 26, 769–778. 179–184. Álvarez, I., Wendel, J.F., 2003. Ribosomal ITS sequences and plant phylogenetic Joly, S., McLenachan, P.A., Lockhart, P.J., 2009. A statistical approach for inference. Mol. Phylogenet. Evol. 29, 417–434. distinguishing hybridization and incomplete lineage sorting. Am. Nat. 174, Arnold, M.L., 2016. Anderson’s and Stebbins’ prophecy comes true: genetic E54–E70. exchange in fluctuating environments. Syst. Bot. 41, 4–16. Kadereit, J.W., 2015. The geography of hybrid speciation in plants. Taxon. http://dx. Avise, J.C., Shapiro, J.F., Daniel, S.W., Aquadro, C.F., Lansman, R.A., 1983. doi.org/10.12705/644.I. Mitochondrial DNA differentiation during the speciation process in Kluge, A.G., 1989. A concern for evidence and a phylogenetic hypothesis of Peromyscus. Mol. Biol. Evol. 1, 38–56. relationships among Epicrates (Boidae, Serpentes). Syst. Zool. 38, 7–25. Baldwin, B.G., Sanderson, M.J., Porter, J.M., Woiciechowski, M.F., Campbell, C.S., Kluge, A.G., Farris, S.J., 1969. Quantitative phyletics and the evolution of anurans. Donoghue, M.J., 1995. The ITS region of nuclear ribosomal DNA: a valuable Syst. Zool. 18, 1–32. source of evidence on angiosperm phylogeny. Ann. Mo. Bot. Gard. 82, 247–277. Kubatko, L.S., Degnan, J.H., 2007. Inconsistency of phylogenetic estimates from Baldwin, B.G., Markos, S., 1998. Phylogenetic utility of the external transcribed concatenated data under coalescence. Syst. Biol. 56, 17–24. spacer (ETS) of 18S–26S rDNA: congruence of ETS and ITS trees of Calycadenia. Kuppler, A.L.M., Fagúndezb, J., Bellstedt, D.U., Oliver, E.G.H., Léon, J., Pirie, M.D., Mol. Phylogenet. Evol. 10, 449–463. 2015. Testing reticulate versus coalescent origins of Erica lusitanica using a Baum, D.A., Small, R.L., Wendel, J.F., 1998. Biogeography and floral evolution of species phylogeny of the northern heathers (Ericeae, Ericaceae). Mol. Phylog. baobabs Adansonia, Bombacaceae as inferred from multiple sata sets. Syst. Biol. Evol. 88, 121–131. 47, 181–207. Kvist, L.P., Skog, L.E., 1993. The genus Columnea (Gesneriaceae) in Ecuador. Bentley, J., Verboom, G.A., Bergh, N.G., 2014. Erosive processes after tectonic uplift Allertonia 6, 327–400. stimulate vicariant and adaptive speciation: evolution in an Afrotemperate- Lee, R.E., Sherk, L.C., 1963. Thirteen new Columnea hybrids from Cornell University. endemic paper daisy genus. BMC Evol. Biol. 14, 27. Gloxinian 13 (114–120), 172–176. Buckley, T.R., Cordeiro, M., Marshall, D.C., Simon, C., 2006. Differentiating between Levin, D.A., 1978. The origin of isolating mechanisms in flowering plants. Evol. Biol. hypotheses of lineage sorting and ontrogression in New Zeeland alpine cicadas 11, 185–317. (Maoricicada Dugdale). Syst. Biol. 55, 411–425. Levin, D.A., 2000. The Origin, Expansion, and Demise of Plant Species. Oxford Byrne, R., Morley, B., 1976. Hybridization studies in Columnea L. (Gesneriaceae). 2. University Press, New York. The C. querceti complex. Bot. J. Linn. Soc. 72, 199–210. Lexer, C., Widmer, A., 2008. The genic view of plant speciation: recent progress and Clark, J.L., Herendeen, P.S., Skog, L.E., Zimmer, E.A., 2006. Phylogenetic relationships emerging questions. Philos. Trans. Roy. Soc. B 363, 3023–3036. and generic boundaries in the tribe Episcieae (Gesneriaceae) inferred from Linder, C.R., Rieseberg, L.H., 2004. Reconstructing patterns of reticulate evolution in nuclear, chloroplast, and morphological data. Taxon 55, 313–336. plants. Am. J. Bot. 91, 1700–1708. Clark, J.L., Funke, M.M., Duffy, A.M., Smith, J.F., 2012. Phylogeny of a Neotropical Lumaret, R., Jabbour-Zahab, R., 2009. Ancient and current gene flow between two clade in the Gesneriaceae: more tales of convergent evolution. Int. J. Plant Sci. distantly related Mediterranean oak species, Quercus ruber and Q. ilex. Ann. Bot. 173, 894–916. 104, 725–736. Clay, D.L., Novak, S.J., Serpe, M.D., Tank, D.C., Smith, J.F., 2012. Homoploid hybrid Maddison, W.P., 1997. Gene trees in species trees. Syst. Biol. 46, 523–536. speciation in a rare endemic Castilleja from Idaho (Castilleja christii, Marín-Gómez, O.H., 2014. Ecología de la polinización y mecanismos de coexistencia Orobanchaceae). Am. J. Bot. 99, 1976–1990. del ensamble de Columnea (Gesneriaceae) en la reserva natural Ñambí, Nariño- Comes, H.P., Abbott, R.J., 2001. Molecular phylogeography, reticulation, and lineage Colombia Msc Thesis. Universidad Nacional De Colombia, Bogotá. sorting in Mediterranean Senecio Sect. Senecio (Asteraceae). Evolution 55, 1943– Marín-Gómez, O.H., Amaya-Márquez, M., 2015. Diversidad, densidad poblacional y 1962. distribución espacial de Columnea (Gesneriaceae) en la Reserva Natural Río Coyne, J.A., Orr, H.A., 2004. Speciation. Sinauer, Sunderland, MA. Ñambí, Nariño. Colombia. Rev. Acad. Colomb. Cienc. Ex. Fis. Nat. 39, 218–227. Wendel, J.F., Cronn, R.C., 2003. Polyploidy and the evolutionary history of cotton. Martinsen, G.D., Whitham, T.G., Turek, R.J., Keim, P., 2001. Hybrid populations Adv. Agron. 78, 139–186. selectively filter gene introgression between species. Evolution 55, 1325–1335. Cummings, M.P., Neel, M.C., Shaw, K.L., 2008. A genealogical approach to Mason-Gamer, R.J., Kellogg, E.A., 1996. Chloroplast DNA analysis of the quantifying lineage divergence. Evolution 62, 2411–2422. monogenomic Triticeae: phylogenetic implications and genome-specific Darwin, C., 1859. On the Origin of Species by Means of Natural Selection. Murray, markers of special interest. In: Jauhar, P. (Ed.), Methods of Genome Analysis London. in Plants. CRC Press, Boca Raton, pp. 301–325. Denduangboripant, J., Cronk, Q.C.B., 2000. High intraindividual variation in internal Maureira-Butler, I.J., Pfeil, B.E., Muangprom, A., Osborn, T.C., Doyle, J.J., 2008. The transcibed spacer sequences in Aeschynanthus (Gesneriaceae): implications for reticulate history of Medicago (Fabaceae). Syst. Biol. 57, 466–482. phylogenetics. Proc. Roy. Soc. B 267, 1407–1415. 240 J.F. Smith et al. / Molecular Phylogenetics and Evolution 106 (2017) 228–240
Mayr, E., 1942. Systematics and the Origin of Species. Columbia University Press, Seelanan, T., Schnabel, A., Wendel, J.F., 1997. Congruence and consensus in the New York. cotton tribe (Malvaceae). Syst. Bot. 22, 275–288. Miller, M.A., Pfeiffer, W., Schwartz, T., 2010. Creating the CIPRES science gateway for Shaw, J., Lickey, E.B., Schilling, E.E., Small, R.L., 2007. Comparison of whole inference of large phylogenetic trees. In: Proceedings of the Gateway chloroplast genome sequences to choose noncoding regions for phylogenetic Computing Environments Workshop (GCE), 14 Nov. 2010, New Orleans, LA, studies in angiosperms: the tortoise and the hare III. Am. J. Bot. 94, 275–288. pp. 1–8. Sherk, L.C., Lee, R.E., 1967. Interspecific hybridization in the genus Columnea Moore Jr., H.E., 1954. Article 17. Some cultivated Gesneriaceae and hybrids. Gentes (Gesneriaceae). Baileya 15, 89–96. Herbarum 8, 375–403. Shimodaira, H., 2002. An approximately unbiased test of phylogenetic tree Morley, B.D., 1971. A hybrid swarm between two hummingbird-pollinated species selection. Syst. Biol. 51, 492–508. of Columnea (Gesneriaceae) in Jamaica. J. Linn. Soc. Bot. 64, 81–96. Shimodaira, M., Hasegawa, H., 1999. Multiple comparisons of log-likelihoods with Morley, B.D., 1975. More Columnea hybrids. Gloxinian 25, 5–7. applications to phylogenetic inference. Mol. Biol. Evol. 16, 1114–1116. Morley, B.D., 1976. Hybridization studies in Columnea L. (Gesneriaceae). 1. Jamaican Skog, L.E., 1978. Flora of Panama, Family 175. Gesneriaceae. Ann. Missouri Bot. Gard. species. Bot. J. Linn. Soc. 72, 191–198. 65, 783–996. Morrison, D.A., 2009. Using data-display networks for exploratory data analysis in Small, R.L., Ryburn, J.A., Cronn, R.C., Seelanan, T., Wendel, J.F., 1998. The tortoise and phylogenetic studies. Mol. Biol. Evol. 27, 1044–1057. the hare: choosing between noncoding plastome and nuclear Adh sequences for Müller, K., 2004. PRAP – computation of Bremer support for large datasets. Mol. phylogeny reconstruction in a recently diverged plant group. Am. J. Bot. 85, Phylogenet. Evol. 31, 780–782. 1301–1315. Müller, K., Quandt, D., Müller, J., Neinhuis, C., 2005. PhyDE 0.9971: Phylogenetic Smith, B.T., McCormack, J.E., Cuervo, A.M., Hickerson, M.J., Aleixo, A., Cadena, C.D., Data Editor.